1 .Kvalitative intervju Aslaug G. Almås
Intervju eller dialog? Dialog mer likestilt Intervju: Annet maktforhold. Intervjuer bestemmer/tolker spørsmålene.
Sju stadium i intervju undersøk. Tematisering: Mål med undersøkelse. Hvorfor og hva må komme før hvordan Planlegging: Gå gjennom alle steg før start Øving: Aldri godt nok forberedt. Bruk intervju guide, gjør prøveintervju, lær av andre. Transkribering: Fra muntlig tale til skriftlig tekst. Gjør det selv. Best å transkribere alt, ikke bare å velge subjektivt. Analysering: Forskers kompetanse er avgjørende. Viktig å kunne research topic. NVIVO: programvare som kan bidra i analysen Analyse metoder: Meningskategorisering (+/-5), meningsfortetting, meningsstrukturering gjennom narrativ, meningstolkning, og ad-hoc metoder for meningsgenerering. Dypere analysering: Tolkingskontekster og valideringsfellesskap Verifisering: Generaliserbarhet, reliabilitet og validitet.
Rapportering: Lesbar, ofte kjedelig, mange sitat. Tenk på rapportering før man gjør intervju.
Analyse av Narrativer: En kvalitiativ tilnærming til forskningsfeltet. : Anne K. Rønsen En historie som beskriver et fenomen.
Tre dimensjoner: Skal observeres simultant
Temporalitet – man er i en prosess -
Kollektivitet – personlige og sosiale dimensjonen. Relasjoner til andre.
Sted – fysisk sted, rekkefølge av ulike steder.
Design av N forskingsløp Rettferdigjøre metodevalget (why?) . Hvilket fenomen dreier forskingen seg om. Hensiktsmessige metoder? Analyse og tolkning.
Observasjon, refleksjonstekster, transkripsjon av video og lydopptak. Observasjon, samtale.
Retninger i narrativ analyse:
Rogan og de Kock
1. Performative methods
2. Structural methods
3. Litterary methods.
Polkinghorne:
Analysis of narrative
Webster:
Kritiske hendelser blir ofte referert tilbake til
Using narrative Inquiery
G. Salomon
What is worth studying and what isn't?
A triviel question:
If i teach them...will they learn?
- A good research question needs to be real question about something not well known.
Unresearchable questions. (non-scientific)
- What is the best ICT to teach history?
- Cannot deal with 15 alternatives to be compared. Needs to deal with the essence of something, not private cases.
Studying something that is not an essential characteristic but a private case and thus would not allow any generalisations.
e.g. why dont teachers se program x?
For every X there are many more x's. A good research question would ask what makes teachers avoid a certain class of tools.
Studying the wrong independent variable:
e.g. studying the effects of a tool while in fact it is the team process that makes the difference in learning.
A good research question is based on prior identification of what really would make the difference.
Defining a research question without knowing the relevant literature
e.g. expecting cognitive transfer effects by an ICT program without being aware of 90 years of research on it.
A good research question stands on the shoulders of past research; it continues from there.
Studyng the effects of an ICT based activity on general abilities without a rationale as to why this miracle should happen.
A good research question is based on a convincing literature-based rationale as to why this may happen. (the argument "because I believe so" is no rationale).
Expecting overnight results
e.g. playing a simulation game in a team will improve communication skills.
A good research design assumes that effects take time; nothig happens overnight.
Looking for main effects where interactions should be expected
e.g. Is independent internet search accompanied by student satisfaction?
A good research question would be: who is satisfied with the independent search and who is not? Why?
Intervju eller dialog?
Dialog mer likestilt
Intervju: Annet maktforhold. Intervjuer bestemmer/tolker spørsmålene.
Sju stadium i intervju undersøk.
Tematisering:
Mål med undersøkelse. Hvorfor og hva må komme før hvordan
Planlegging: Gå gjennom alle steg før start
Øving: Aldri godt nok forberedt. Bruk intervju guide, gjør prøveintervju, lær av andre.
Transkribering: Fra muntlig tale til skriftlig tekst. Gjør det selv. Best å transkribere alt, ikke bare å velge subjektivt.
Analysering: Forskers kompetanse er avgjørende. Viktig å kunne research topic. NVIVO: programvare som kan bidra i analysen Analyse metoder: Meningskategorisering (+/-5), meningsfortetting, meningsstrukturering gjennom narrativ, meningstolkning, og ad-hoc metoder for meningsgenerering.
Dypere analysering: Tolkingskontekster og valideringsfellesskap
Verifisering: Generaliserbarhet, reliabilitet og validitet.
Rapportering: Lesbar, ofte kjedelig, mange sitat. Tenk på rapportering før man gjør intervju.
Analyse av Narrativer: En kvalitiativ tilnærming til forskningsfeltet. : Anne K. Rønsen
En historie som beskriver et fenomen.
Tre dimensjoner: Skal observeres simultant
Temporalitet – man er i en prosess -
Kollektivitet – personlige og sosiale dimensjonen. Relasjoner til andre.
Sted – fysisk sted, rekkefølge av ulike steder.
Design av N forskingsløp
Rettferdigjøre metodevalget (why?) . Hvilket fenomen dreier forskingen seg om. Hensiktsmessige metoder? Analyse og tolkning.
Observasjon, refleksjonstekster, transkripsjon av video og lydopptak. Observasjon, samtale.
Retninger i narrativ analyse:
Rogan og de Kock
1. Performative methods
2. Structural methods
3. Litterary methods.
Polkinghorne:
Analysis of narrative
Webster:
Kritiske hendelser blir ofte referert tilbake til
Using narrative Inquiery
G. Salomon
What is worth studying and what isn't?
A triviel question:
If i teach them...will they learn?
- A good research question needs to be real question about something not well known.
Unresearchable questions. (non-scientific)
- What is the best ICT to teach history?
- Cannot deal with 15 alternatives to be compared. Needs to deal with the essence of something, not private cases.
Studying something that is not an essential characteristic but a private case and thus would not allow any generalisations.
e.g. why dont teachers se program x?
For every X there are many more x's. A good research question would ask what makes teachers avoid a certain class of tools.
Studying the wrong independent variable:
e.g. studying the effects of a tool while in fact it is the team process that makes the difference in learning.
A good research question is based on prior identification of what really would make the difference.
Defining a research question without knowing the relevant literature
e.g. expecting cognitive transfer effects by an ICT program without being aware of 90 years of research on it.
A good research question stands on the shoulders of past research; it continues from there.
Studyng the effects of an ICT based activity on general abilities without a rationale as to why this miracle should happen.
A good research question is based on a convincing literature-based rationale as to why this may happen. (the argument "because I believe so" is no rationale).
Expecting overnight results
e.g. playing a simulation game in a team will improve communication skills.
A good research design assumes that effects take time; nothig happens overnight.
Looking for main effects where interactions should be expected
e.g. Is independent internet search accompanied by student satisfaction?
A good research question would be: who is satisfied with the independent search and who is not? Why?
Most of our main effects are likely to be interactions with:
Relevant individual differences
Particular contents: learning mathematics is not like learning history
(am i looking for main effects or merely interactions?)
Specific motivations: need achievement is not the same as curiosity
Different context for study: Home is not a classroom.
(Aptutude treatment, interaction. ATI)
http://tip.psychology.org/cronbach.html
http://www.emtech.net/learning_theories.htm
http://eric.ed.gov/ERICWebPortal/Home.portal?_nfpb=true&ERICExtSearch_SearchType_0=eric_metadata&_pageLabel=ERICSearchResult&newSearch=true&searchtype=keyword&ERICExtSearch_SearchValue_0=EJ369291